Profile-reward recipe: generate 640 (batched) → ASR-filter+trim → SIDON → keep the top 20% by a per-character quality+profile reward → warm-start FT 3 epochs. Below: the quality range of the selected training set, then example generations after each epoch.
dark_iron_dwarf
Characteristic dimensions (auto, from its own voice profile): ▲ S_STRY ▲ AROU ▲ STRU ▲ S_MONO ▼ EXPL ▼ S_RANT ▼ S_ASMR ▼ S_TECH · emo: Concentration Interest Emotional_Numbness
Training set = top-20% by reward = z(genu)+z(blend)+z(quality)−z(WER)+0.7·z(profile-match). Kept 125/627.
Best 5 of the kept top-20%
R 7.59 · genu 0.27 · blend 0.48 · qual 0.43 · WER 0.00
R 7.14 · genu 0.22 · blend 0.49 · qual 0.49 · WER 0.00
R 6.57 · genu 0.17 · blend 0.35 · qual 0.53 · WER 0.00
R 5.94 · genu 0.09 · blend 0.43 · qual 0.56 · WER 0.00
R 5.73 · genu 0.27 · blend 0.15 · qual 0.64 · WER 0.00
Weakest 5 still inside the top-20% (the selection boundary)
R 1.27 · genu 0.03 · blend 0.04 · qual 0.64 · WER 0.00
R 1.26 · genu 0.00 · blend 0.01 · qual 0.70 · WER 0.00
R 1.26 · genu 0.07 · blend 0.00 · qual 0.60 · WER 0.00
R 1.26 · genu 0.08 · blend 0.25 · qual 0.33 · WER 0.00
R 1.26 · genu 0.05 · blend 0.02 · qual 0.66 · WER 0.00
Fantasy line — refined ep1 / ep2 / ep3
“The old road winds past the ruined tower where the last dragon knights once kept their watch.”
Sci-fi line — refined ep1 / ep2 / ep3
“The station drifted in low orbit while the reactor cooled and the scanners swept the empty dark.”
Neutral line — refined ep1 / ep2 / ep3
“The meeting is scheduled for three o'clock and the report is on the desk by the printer.”